We have developed multiple GEM and GEM-derived allograft (GDA) cancer models for preclinical biomarker and therapeutic development. In addition, responding to growing preclinical needs and supporting collaborative activities initiated by CCR clinical investigators, CAPR specialists also recently embarked on a mission to establish several patient-derived xenograft models (PDX), to be employed for therapeutic discovery in oligo-proliferative lung cancer cases. On the path of building a state-of-art portfolio of preclinical models, CAPR major accomplishments include: 1. Generation of multiple driver-specific GEM and GDA metastatic human-relevant SEOC models along with multiple cell cultures for each Inducible models that have been shown to develop SEOC that resembles the human disease in both molecular and biological properties. Models that retain wild type BRCA1 and BRCA2 and models that are deficient in BRCA1 or 2 were established. Models were validated to respond to PARP inhibition with outcomes similar to those in clinical cases and are currently employed in several collaborative studies (also in collaboration with major Pharma company) pursuing validation of combinatorial strategies to attack SEOC malignancy. In addition, a collection of novel ovarian models is being assembled in which oviduct epithelial cells are genetically programmed to give rise to SEOC carcinogenesis representing a current paradigm that the secretory cells of human fallopian tubes represent the cell of origin for SEOC cancer. 2. Generation of human-relevant GEM and GDA primary GBM models and cell lines. Tractable intracranial transplant GBM models were developed and robust therapeutic evaluation workflows were established. The combination of MEK and PI3 kinase inhibition has been demonstrated to inhibit tumor growth and extend survival, whereas inhibition of either alone proved less effective. Development of a human EGFR-driven erlotinib resistant focally evolving lung adenocarcinoma (LA) model All previously established models of this type develop diffuse adenocarcinoma lesions, making it difficult to follow disease for enrollment into drug studies and to monitor and quantify efficacy. We developed a model wherein the erlotinib resistant human EGFRL858R;T790M drives focal development of LA. The model has been validated for increased accuracy of tumor development and efficacy evaluation. 3. Multiple early passage cell cultures have been established from pancreatic cancerous lesions that developed in the KPC GEM model (established by Drs. Tuveson and Hingorani). These cells, when transplanted orthotopically into the pancreas of recipient immunocompetent mice, develop metastatic pancreatic adenocarcinomas with complete set of characteristics typical for the parental tumor. These models have the highest metastatic rate of any GEM PDAC model reported thus far. 4. We have completed the construction and extensively validated three novel GEM strains for the most common p53 missense mutations identified in human cancer cases (R172H, R270H, and R270C). Pancreatic Ductal Adenocarcinoma (PDAC) models harboring these p53 alleles have been generated by inter-crossing these alleles with mice harboring the inducible RasG12D allele and PDX-Cre activating transgene. Mice harboring the common p53-172 mutant have been characterized and shown to develop PDAC with properties similar to the KPC model. 5. As an exploratory technology adoption effort, CRISPR/Cas9 system was optimized to introduce multiple oncogenic mutation in lung and brain cells using in vivo infection by recombinant Adeno-Associated Viral (rAAV) Vectors. Intra-tracheal infusion of rAAV reagents engineered to introduce a common spectrum of pro-oncogenic mutations detected in clinical non-small cell lung cancer cases (s.u. p53, Lkb1, and PTEN deletions, Kras-G12D, -D12V, G13D missense mutations, etc.) led to development of pulmonary adenocarcinomas with 5-7-month latency. This technology is being optimized to enable establishing cohorts of experimental animals using such somatic genome editing strategy to introduce oncogenic events in adult tissues of interest, including (but not limited to) brain, lung and pancreas. 6. Via homologous recombination in ES cells, CAPR model development experts have constructed and validated a CreERT expressing tamoxifen-inducible 'deletor' mouse strain in which an expression of tamoxifen-dependent CreERT2 recombinase is controlled by the endogenous locus for dopachrome tauto-isomerase 1 (Dct1), known to be active in neural crest derived cells, such as melanocyte progenitor cells called melanoblasts. Upon extensive molecular validation of this new strain, we have intercrossed the Dct-1-CreERT2 allele with two conditional strains carrying Cre-inducible alleles for Braf-V600E mutation (very common in clinical metastatic melanoma cases) and PTEN (also reported lost in melanoma tumors). Tri-allelic animals displayed robust carcinogenesis upon induction with tamoxifen, forming rapidly progressing cutaneous neoplasms. Intriguingly, histopathologic and biomarker assessment suggested these tumors to represent peripheral nerve sheath tumors (or PNST) variety of cutaneous malignancies, that can be explained by the activity of Dct1 promotor in neural crest cells that also give raise to Schwann cells of the peripheral nervous system. CAPR scientists are now collaborating with specialists in the CCR Rare Tumor Initiative to conduct a more detailed description of this model, but also identify alternative allelic combination(s) and/or tumor induction strategies to obtain animals capable of developing melanomas. 7. In collaboration with Dr. Sheue-yann Cheng;s lab at the Pediatric Oncology Branch, we are developing a new Tpo-CreERT2 allele that will express CreERT recombinase in thyroid gland C-cells. Once established and validated, this strain will be used to assemble a genetic model for medullary thyroid carcinoma, also using conditional p53 missense mutation alleles generated at CAPR and shared with Dr. Cheng. In addition to the aforementioned achievements in cancer modeling, two integrated databases were developed tor managing (1) efficient GEM breeding and maintenance workflows, and (2) preclinical therapeutic evaluation, workflow execution, and data collection in genetic model for pancreatic cancer. 8. In collaboration with Drs. Carol Thiele and Jack Shern, several pdx rhabdomyosarcoma and GEM allograft models for neuroblastoma have been established and characterized for subsequent preclinical testing of JQ1 BET bromodomain-containing protein inhibitors and their combination with common chemotherapeutics, such as irinotecan.